An enhanced Automatic Collision Notification System enables three quarters (75.9 percent) of injured occupants to be correctly identified as seriously injured, by using only data automatically collected and transmitted by the vehicles.
Since 2006, nearly 14,000 BMW crashes have occurred in the US involving vehicles equipped with ACN or enhanced ACN technology. Of these, 70 percent of occupants indicate no injury to the TSP (Telematics Service Provider) operators, 20 percent indicate they are injured in some way and require help while 10 percent provide no verbal response to the TSP call-taker. An investigation of a subsample of crashes occurring in Florida suggests that no hospital transport was necessary for 81 percent of the calls where no voice response occurred. Although the majority of these cases require no further care, 19 percent of the no voice population was subsequently transported to a hospital or trauma center for additional care. This population of occupants could benefit from an automatic call for help to a Public Services Answering Point (PSAP - commonly known as 911) that includes an estimate of the likelihood of serious injuries.
To assist in identifying crashes with incapacitating injuries, the William Lehman Injury Research Center (WLIRC) in Miami, Florida and BMW have pioneered the development of an algorithm called URGENCY. This algorithm is based on US national crash statistics and BMW internal data. The injury prediction by URGENCY permits the transmission of the earliest and best information to the PSAP. We report early observations of injury severity and location for enhanced ACN equipped vehicle crashes occurring in the US and Germany.
Seriously injured vehicle occupants are defined as those who have sustained one or more injuries with an Abbreviated Injury Severity (AIS) Score of 3 or higher (includes
AIS 3, AIS 4, AIS 5, AIS 6 and fatally injured). This group is referred to as MAIS3+ injured and includes those who need immediate medical attention due to potentially life threatening injuries.
This study evaluated how well a real-time computer model (using an algorithm called URGENCY) was able to estimate the risk of serious injuries based on crash conditions (that were communicated to the call center). The overall predictive accuracy of the model suggests that 75.9 percent of injured occupants would be correctly identified using data automatically collected and transmitted by vehicles alone. In other words, an automatic call for help indicating serious injury would most likely be made for three out of four MAIS3+ injured occupants even if their crash was not observed by somebody on scene or if occupants were unable to place a call themselves. When URGENCY estimates are used in combination with verbal information gathered by the TSP or 911, occupants in need of medical attention would rarely be missed. A third opportunity to assess injury severity exists before hospital transport once EMS has arrived on scene. Rapid and accurate identification of seriously injured vehicle occupants is critical to quickly despatching the appropriate resources and quickly getting the injured occupants to the appropriate medical facilities, which is essential for saving lives.
Author: Stefan Rauscher, Georg Messner, Peter Baur, Jeffrey Augenstein, Kennerly Digges, Elana Perdeck, George Bahouth, Oliver Pieske
Published By: National Highway Traffic Safety Adminstration (NHTSA)
Paper Presented at the 21st International Technical Conference on the Enhanced Safety of Vehicles
Source Date: June 2009URL: http://www-nrd.nhtsa.dot.gov/pdf/esv/esv21/09-0049.pdf
Average User Rating
Related Metropolitan Integration Links
Typical Deployment Locations
automated collision notification, ACN